jialicheng/unlearn_samsum_t5-base_neggrad_4_42

samsum_42

This model is a fine-tuned version of google/t5-v1_1-base on the samsum dataset. It achieves the following results on the evaluation set:

• Loss: 1.7890
• Rouge1: 40.3856
• Rouge2: 20.2717
• Rougel: 35.4629
• Rougelsum: 37.3525
• Gen Len: 11.9010
• Test Rougel: 35.4629
• Df Rougel: 35.4391
• Unlearn Overall Rougel: 0.5119
• Unlearn Time: 413.9972

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-05
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len	Overall Rougel	Unlearn Overall Rougel	Time
No log	1.0	19	1.5087	48.165	24.0364	41.5145	44.5273	20.0330	0.0250	0.0250	-1
No log	2.0	38	1.5518	45.1367	22.6241	39.4987	41.5893	15.9988	0.1158	0.1158	-1
No log	3.0	57	1.6497	42.676	21.8348	37.1472	39.346	13.2311	0.5080	0.5080	-1
No log	4.0	76	1.7476	40.7242	20.605	35.8234	37.7693	12.0526	0.5084	0.5084	-1
No log	5.0	95	1.7890	40.3856	20.2717	35.4391	37.3525	11.9010	0.5119	0.5119	-1

Framework versions

• Transformers 4.39.3
• Pytorch 2.3.0+cu121
• Datasets 2.19.0
• Tokenizers 0.15.2